Cutaneous malignant melanoma (MM) is basically an uncontrolled overgrowth of neoplastic melanocytes. At some stage of its progression, the neoplasm exhibits a high metastatic potential. It proves to be resistant to drug-induced apoptosis, which is believed to underlie the resistance of MM to conventional chemotherapy and radiotherapy (1,2). Various interactions exist between MM cells and other biological systems, including immune cells, vascularity, contiguous stromal cells and the dermal extracellular matrix (ECM). Certain aspects of MM-stroma interactions are thought to be associated with disease prognosis (3). In addition, environmental influences, including ultraviolet (UV) light, are probably responsible for MM initiation and may support its progression along with the intervention of diverse autocrine and paracrine factors (4). In particular, a number of growth factors and specific enzymes are released in the MM microenvironment (5–7).

The participation of the host in the ‘cancer micro-ecosystem’ basically involves the microvasculature, stromal cells and specific immune reactions (8–10). Angiogenesis is a typical host-mediated response to many cancers. It appears crucial for cancer progression, as blood vessels deliver nutrients and oxygen to neoplastic cells (11). Furthermore, the microvasculature likely allows communication between the primary MM and its metastases. Pro-angiogenic molecules originate from cancer cells as well as from the stroma. The relative contribution of both compartments is likely to change with MM type and site, and is balanced by other factors as well (11,12).

Cross-talk between MM and stromal cells may be mediated through direct heterotypic cell-cell contacts, adhesion molecules, signaling factors, and other secreted molecules consisting of growth factors, cytokines, chemokines, ECM proteins, proteinases, proteinase inhibitors and lipid products (13). Conceptually, the MM microenvironment is crucial for the maintenance of cellular functions and tissue integrity, suggesting that a cancer-induced change in the ECM may contribute to cancer invasion (14). Any alteration in the MM stroma may be due to an imbalance in the cytokine profile, resulting from oncogenic changes in the cancer cells. In particular, experimental animal models have demonstrated that cancer invasion is stimulated by the wound-healing stroma (15).

Both stromal cells and the ECM located beneath primary MM lesions are therefore likely involved in the process of invasion of the neoplasm and in the early dissemination of micrometastases associated or not with neoangiogenesis (2,10,16–19). These characteristics are possibly associated with phenotypic changes in the stromal cells in the MM vicinity. In recent years, tumor growth regulation by ECM components has been one of the main topics of neoplastic biology research.

This study is a review of current peer-reviewed publications admixed with personal original findings from a series of 400 MM cases with a thickness ranging between 0.4 and 1.0 mm (median 0.83) that were retrieved from our files. The microscopic diagnosis was previously established by a group of three dermatopathologists. Immunohistochemistry was performed as previously described (20–23). In short, samples were fixed in buffered formalin and embedded in paraffin. A series of 6-μm sections were prepared for immunohistochemistry. The avidin-biotin peroxidase method was used with the antibodies listed in Table I. After 1 h of incubation with each primary antibody, the slides were washed in Tris-buffered saline (TBS) and incubated for 30 min with the secondary antibody (biotinylated swine anti-rabbit, 1:300; Dakopatts Glostrup, Denmark). Slides were rinsed in TBS and covered by the EnVision (Dakopatts) polymer-based revelation system. After TBS washings, Fast Red (Dakopatts) was used as the chromogen substrate. The final steps consisted of counter-staining with Mayer's hemalum and mounting in glycerin mounting medium (Dakopatts). Negative immunohistochemical controls were performed by omitting or substituting the primary and the secondary antibodies from the laboratory procedure.

A typical biological feature of human MM is the tremendous impact of the primary lesion thickness on prognosis. Primary MM <1 mm in thickness is associated with a high cure rate, sharply contrasting with thicker lesions associated with poorer prognosis. The apparent breakpoint beyond an ∼1-mm thickness is a discouraging factor in disease outcome. One possible reason appears to be linked to vascularization patterns of MM (11,12). Such an anatomic argument is persuasive, but it is by no means the only one.

MM cells apparently fail to form metastases unless they present the genotypic and phenotypic information allowing them to effectively migrate in the ECM, intravasate, extravasate, cross interstitial basement membranes and proliferate in distant tissue sites. These characteristics are expressed by variant subpopulations of metastatically competent MM cells present in primary neoplasms (2). These subpopulations probably acquire a growth advantage at the primary site over time, so that they become a dominant proliferating population. At this stage, the MM truly expresses overt malignancy. As a result of this process of clonal dominance of metastatically competent cells, it is possible that most thin primary MM lesions contain very few, if any, metastatically competent cells, whereas thicker MM lesions may contain significant proportions of such cells. It is possible that the stromal microenvironment plays a role in such a shift in the biological profile of MM (24).

Upon standard histopathological examination, stromal cells appear normal underneath primary MM lesions when partial regression is not operative. However, their differentiation as revealed by immunohistochemistry appears altered when compared to the surrounding skin. In particular, phenotypic changes are noted when identifying the transglutaminase Factor XIIIa, α (IV) collagen chains, as well as elafin and versican. It is possible, although not yet proven, that transforming growth factor (TGF)-β1 and platelet-derived growth factor (PDGF) may play a role in the alteration of the stromal host compartment in MM.

This review highlights the existence of a distinct region of the dermis adjacent to the base of a primary MM lesion. Stromal cells exhibit particular phenotypic features suggesting altered functionality. The involved territory appears to be conducive to micrometastatic spread. Some of these cells survive singly, and due to their manner of migration to other organs may represent MM stem cells.

In addition to the importance of MM vascularization for tumor growth, invasiveness and metastatic spread (10–12,19,59), numerous other roles are ascribed to the tumoral stroma. This structure is involved in a constant remodeling following degradation and repair of the ECM. Notably, immunohistochemistry highlights the direct implication of MM cells in the synthesis and/or storage of certain ECM molecular components.

The immunohistochemical characterization of MM cells is important (20,22,60–64), yet should be extended to the peritumoral stroma, including the microvasculature (10–12,59) and other ECM components. A comprehensive mapping of MM immunohistochemical characteristics should aid in identifying relevant targeted therapies (63–66).

Inflammatory cells and immunocytes represent another class of host cells that are regulated by the balance of cytokines. They perform counter-current invasion, from the circulation into the tumor, and provide routes for MM cell invasion. It is important for our understanding of MM stroma turnover that tumor-infiltrating leukocytes produce proteinases.

Interaction between MM and its stroma is evident during the invasive and metastatic stages of disease progression. Stromal cells are known to secrete metalloproteinases and their inhibitors, growth factors, the scatter factor/hepatocyte growth factor and other factors, as well as participate in the growth and mobility of MM cells. In addition, other molecules are synthesized and overexpressed by stromal cells and/or MM cells. Immunohistochemistry has identified Factor XIII-a, α1 and α5 (IV) collagen chains, versican, elafin and lysozyme. These possibly influence the migration of MM cells, including their stem cells.

While MM cell motility cannot be directly assessed, there is circumstantial evidence indicating that motility is essential to MM progression and possibly of prognostic significance. Apart from the secretion and activation of enzymes altering the ECM, a variety of stromal alterations occur following overexpression of diverse ECM components. Molecular morphology yields evidence suggesting that the MM stroma plays an integral role in MM. Although much remains to be determined, the findings as described in the present review may have diagnostic and prognostic significance, which warrant further investigation.